pith. sign in

arxiv: 2202.10042 · v1 · pith:MPIM7USAnew · submitted 2022-02-21 · 🧮 math.OC

Fast Sinkhorn I: An O(N) algorithm for the Wasserstein-1 metric

classification 🧮 math.OC
keywords algorithmsinkhornmetriccomputationalefficientmultiplicationswassersteinaccuracy
0
0 comments X
read the original abstract

The Wasserstein metric is broadly used in optimal transport for comparing two probabilistic distributions, with successful applications in various fields such as machine learning, signal processing, seismic inversion, etc. Nevertheless, the high computational complexity is an obstacle for its practical applications. The Sinkhorn algorithm, one of the main methods in computing the Wasserstein metric, solves an entropy regularized minimizing problem, which allows arbitrary approximations to the Wasserstein metric with O(N^2) computational cost. However, higher accuracy of its numerical approximation requires more Sinkhorn iterations with repeated matrix-vector multiplications, which is still unaffordable. In this work, we propose an efficient implementation of the Sinkhorn algorithm to calculate the Wasserstein-1 metric with O(N) computational cost, which achieves the optimal theoretical complexity. By utilizing the special structure of Sinkhorn's kernel, the repeated matrix-vector multiplications can be implemented with O(N) times multiplications and additions, using the Qin Jiushao or Horner's method for efficient polynomial evaluation, leading to an efficient algorithm without losing accuracy. In addition, the log-domain stabilization technique, used to stabilize the iterative procedure, can also be applied in this algorithm. Our numerical experiments show that the newly developed algorithm is one to three orders of magnitude faster than the original Sinkhorn algorithm.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Effective dynamics of the Sinkhorn algorithm in the regime of low entropy regularization

    math.OC 2026-07 unverdicted novelty 7.0

    Derives the cold Sinkhorn limiting dynamics as tau approaches zero, proving finite-time convergence to unregularized OT and improved O(tau^{-1}) iteration complexity for dual suboptimality.

  2. Sharp $O(1/k)$ convergence rate for the Sinkhorn algorithm via a local analysis

    math.OC 2026-06 unverdicted novelty 7.0

    Proves sharp O(1/k) rate for Sinkhorn via local bipartite graph analysis of positive-mass edges, bootstrapped from prior almost-sharp global bound.

  3. Importance Sparsification for Sinkhorn Algorithm

    stat.ML 2023-06 unverdicted novelty 7.0

    Spar-Sink applies importance sampling based on upper bounds of transport plans to create a sparse approximation of the Sinkhorn kernel, yielding consistent estimators for regularized OT and UOT with near-linear iterat...